7% of Vehicles, 39% of Complaints

I run a moving company in Garland, Texas. I drive trucks every day, fix them most weekends, and buy used ones for the fleet a few times a year.

Two years ago I started keeping notes on which trucks were trouble and which weren't. Then I realized NHTSA was already keeping those notes — every recall, every owner complaint, every defect investigation, all sitting in a federal database that nobody outside of safety researchers really reads.

So we built a site to read it. Then I had a question. If you take the engines and transmissions everybody in the shop world already knows are problems — the Theta II, the PowerShift, the 6.0 Powerstroke, the lifter-eating GM 5.3 — and you actually count up the owner complaints filed against them, how big is that pile?

Here's what we found.

The headline

7.1% of the vehicles in our database account for 38.5% of the NHTSA owner complaints filed against them.

That's the math. We have 16,825 vehicle model-years in the database, covering 33 manufacturers from 2005 through 2025. Owners filed 768,293 complaints with NHTSA across all of them. Of those complaints, 295,560 — over a third — were filed against just 1,192 vehicles. And those 1,192 vehicles all share something in common: they're equipped with one of 28 engine or transmission families that have documented systemic defects.

The concentration ratio works out to 5.42x. A vehicle with one of these 28 engines or transmissions generates, on average, more than five times the complaint volume of a typical US-market vehicle.

That's not a vague claim. It's the count.

The 28 platforms

Ranked by total NHTSA owner complaints across the affected vehicles in our database.

Full per-platform data — complaint patterns by year, recall severity, NHTSA campaign IDs, and editorial commentary — lives on the engine families directory and transmission families directory. Click any one of these and you get the full breakdown.

What the numbers tell you

I spent some time staring at this list before writing anything. Five things stood out.

1. The Ford 6F35 is the worst single platform in our database.

37,679 owner complaints across 63 vehicle model-years. That's about 600 complaints per vehicle on average — roughly four times the typical complaint rate for a vehicle that's been on the road. Ford put this transmission in Escapes, Fusions, MKZs, and a few other midsize products from 2009 onward. It got a class action. It's still in production. The PowerShift DPS6 down at #20 with 8,469 complaints gets all the press, but the 6F35 quietly accumulated four times that volume.

2. Recall counts and complaint counts measure different things.

Look at the Mercedes M272/M273 at #19. 8,664 complaints. Zero active recalls. Mercedes settled a class action on the balance shaft gear and offered limited warranty extensions, but nothing in the formal NHTSA recall database reflects what owners are actually dealing with on these engines.

Or look at the GM 5.3L AFM at #22. 7,680 complaints. Two active recalls. A class action settled regarding oil consumption. The recall count tells you what the manufacturer admitted; the complaint count tells you what owners are filing about anyway.

If you only look at recall counts, you miss most of the picture.

3. Korean engines hit harder than their fleet share would predict.

Hyundai and Kia together represent roughly 9% of the US fleet. Two of their engines (Theta II 2.4L at #2 and Theta II 2.0T at #6) account for over 42,000 complaints in this list — more than 14% of the total complaint volume across the 28 platforms. Their share of the problem is roughly 1.6x their share of the fleet.

This isn't a knock on the brands. The 2014-2019 Theta II rod bearing issue had a class-action settlement and a real engineering fix. But the data says owners filed complaints at a rate that significantly outpaced the brands' market share.

4. CVTs are overrepresented for their fleet share.

Three of the top 10 platforms are CVTs (Subaru Lineartronic at #7, Jatco at #10, and Subaru's CVT issues span multiple year ranges). For a transmission technology that didn't dominate the US fleet through most of this period, CVTs accumulated more complaint volume than you'd expect. Owners file complaints about CVTs more readily than they file about conventional automatics, partly because CVT failures feel different — they vibrate, they hesitate, they fail in ways that don't match what owners expect from a transmission.

5. Direct-injection turbos cluster.

Five of the top 15 platforms are direct-injection turbocharged engines from the 2010s onward (Theta II 2.0T, Ford 3.5 EcoBoost, VW/Audi 2.0T, Ford 1.6 EcoBoost, Honda 1.5T). The shared issues — carbon buildup on intake valves, high-pressure fuel pump failures, oil dilution from cold-start fuel slip — are common to direct-injection turbo design across manufacturers. This isn't a Hyundai problem or a Ford problem. It's a direct-injection-turbo problem, and the manufacturers individually addressed it differently with mixed results.

How we did it

If you're going to cite this analysis or argue with it, here's exactly what we did and where the data comes from.

Source. NHTSA's recallsByVehicle and complaintsByVehicle public APIs. Both are works of the U.S. federal government and are in the public domain. We sync this data weekly to keep it current.

Vehicle universe. 16,825 vehicle model-years covering 33 manufacturers, model years 2005 through 2025.

How we picked the 28 platforms. This is the part of the methodology that involves judgment. We didn't rank engines and transmissions by raw complaint count and pick the top 28 — that would have produced a different list. We picked engines and transmissions that have documented widespread defect patterns: class actions, NHTSA investigations, manufacturer service bulletins, or service-industry consensus that the platform has a systemic problem. Then we measured how much complaint volume those platforms account for. The full curated list with affected vehicle ranges is published on this site, organized by platform: see the engine families directory for all engine platforms with their affected vehicle ranges, and the transmission families directory for transmissions. Each platform page lists every affected year/model combination in our database.

Aggregation. For each platform, we listed every affected vehicle (model-year combinations from production records). We then queried our database for total complaint and recall counts across all those vehicles. Where a vehicle was in both an engine and a transmission family — for example, a 2014 Sonata that's in the Theta II 2.4L list and could also be in a transmission family — we counted it once for the cross-platform total. No double counting.

Concentration ratio. Calculated as (% of total complaints in the platforms) ÷ (% of total vehicles in the platforms). A ratio of 1.0 would mean platform-equipped vehicles complain at exactly the fleet average. We got 5.42x, which means owners of platform-equipped vehicles file complaints at more than five times the average rate.

Data freshness. Numbers in this analysis reflect database state as of May 10, 2026. NHTSA data syncs weekly.

What this doesn't tell you

If you only read one section of this piece, read this one. It's the part most "worst engines" articles skip.

Complaint volume isn't failure rate. A vehicle with 1,000 complaints might not be failing more than one with 500 — it might just have more units on the road, more vocal owners, or more years of accumulated complaints. We don't have production-volume data per vehicle, so we can't compute true per-unit failure rates.

NHTSA underrepresents Tesla. Tesla owners go through Tesla service channels rather than NHTSA at higher rates than other brands. Recent Tesla model years in the database show suspiciously low complaint counts compared to comparable production volumes from other brands. If Tesla doesn't appear on this list, it's not because their cars don't have problems — it's because their owners file complaints differently.

Brand demographics affect filing rates. Mass-market brand owners file with NHTSA more readily than premium brand owners. This biases the data toward overrepresenting domestic and Korean brands. Real failure rates on some BMW and Mercedes platforms are likely higher than complaint volume alone suggests.

Recall count doesn't measure severity. A platform with 65 active recalls might have 64 minor campaigns (paint, trim) and one critical one. We separately classify recalls by severity on individual platform pages, but the headline recall counts in this analysis don't distinguish.

Older model years accumulate more complaints. A 2012 vehicle reported on in 2024 still counts toward the 2012 vehicle's complaint total. Older platforms in this list have had more time to accumulate complaints than newer ones.

The 28 platforms are curated, not exhaustive. We included engines and transmissions with documented widespread defect patterns. Other platforms with smaller but real issues aren't on this list. The 38.5% concentration figure measures specifically these 28 platforms — not all platforms with reliability concerns.

What this means if you're shopping a used vehicle

The practical takeaway is short.

A vehicle with one of these 28 engines or transmissions isn't automatically a bad vehicle. Many of them have well-understood failure modes, known fixes, and ongoing warranty extensions or class-action coverage. With proper maintenance documentation, plenty of them are reasonable buys.

But maintenance documentation matters more than usual on these vehicles. What's been done, when, by whom, with what parts, at what mileage. These questions matter more on a Theta II Sonata or a 6F35 Fusion than they do on a Toyota Corolla with a port-injected 2.5 and a conventional automatic.

If you're shopping a specific used vehicle, the per-platform deep-dives explain the failure modes, the repair costs, and the warranty status:

• Engine families directory — full breakdown on all engine platforms
• Transmission families directory — full breakdown on all transmission platforms

For a per-vehicle reliability profile, browse vehicles by make.

For an honest read on whether an extended warranty makes financial sense on a specific vehicle — including the math on whether you should skip it — the warranty calculator runs the numbers based on real complaint patterns and segment-specific repair costs.

Reproducing this

The full data is open. NHTSA's data is public. Our per-vehicle aggregation is published on individual vehicle pages. Our per-platform aggregation is published on family hub pages. The platform definitions are published on the engine families directory and transmission families directory.

Anybody who wants to verify, extend, or argue with this analysis can. The underlying data is sitting there.